Boyne offers an overview of various empirical studies that focus on the determinants that encourage some local governments to transfer the production of municipal services to external agencies. Boyne stated aim is to answer two questions: 1) To what extent do empirical studies provide an explanation of variations in service contracting? 2) Does the evidence improve our understanding of why different local governments adopt different policies? Boyne limits his examination to those studies examine the following four factors in local government decisions to contract out services: 1) Fiscal Stress, 2) Scale and Market Structure, 3) Public Preferences and 4) Power of Public Employees.
Brief Summary of Boynes Findings
Boyne points out that only three predictor variables used by the studies he examined show any statistically significant singular (positive or negative) impact on the decision to contract out services:
Tax Limits - positive impact on privatization
Income of Population - positive impact on privatization
Public Employees Per Capita - negative impact on privatization
In many other cases variables are shown to have contradictory positive and negative (i.e. tax burden, grants, and population size). Furthermore, R2 results for the models are extremely low, usually below .20 and sometimes as low at .05. This creates additional suspicion of a causal effect of these variables on the likelihood of contracting services.
Boyne notes two general problems with these studies:
These variables are not independent from one another and their combined effect should be examined.
Rather than simply looking whether a service is currently provided by a local government or by a private contractor, it may be more worthwhile to examine whether there as been a recent decision to switch from one to the other. By looking at service provision in a historical context where conscious decisions have obviously been made, a better understanding of the impetus for that decision might be derived.
Summaries of the Studies Boyne Examined
Firstly, Boyne examines ten studies that attempt to determine whether municipalities who are experiencing severe financial pressures are more likely to contract out services. He uses four measures to assess this: 1) Ratio of Local Taxes to Local Incomes because high tax rates are often thought to spur outsourcing to relieve the tax burden; 2) Limits on Local Tax Levels by State Governments because these constraints are thought to force municipalities to better use existing resources; 3) Share of Local Spending funded by Intergovernmental Revenues because increased funding might lessen overall financial pressures, thereby reducing outsourcing. Boyne concludes there is little support for the view that financial stress impacts the decision to contract out services.
Secondly, Boyne examines twelve studies that attempt to determine whether local governments privatize to reap potential benefits of economies of scale and/or by their ability to create a competitive market. They use the following measures: 1) Citizen Population of Municipality a larger cities may have greater purchasing power when negotiating contracts with external venders, 2) Metropolitan Status as the potential for the development of a competitive service market may be different for a local government inside a metropolitan area versus one that is outside. Boyne sees little in the way of a direct effect on the propensity to contract out services coming from either of these two measures and argues that the measures themselves are flawed.
Thirdly, Boyne looks at ten studies that examine the connection between the population characteristics and the levels of contracting out. The variables are: 1) Average Income based on the assumption that high-income groups support contracting out, 2) Percentage of the population that is poor, black, or elderly as these demographic groups are assumed to pressure local governments to maintain direct municipal production of services. There is support for the link between income level and poverty level and the likelihood to outsource services. However, no such connection exists between the proportion of black and elderly populations and the likelihood of outsourcing. Boyne warns that the measure may not necessarily reflect public preference, but simply the ability to pay. He points out flaws in the measures, arguing, that socioeconomic characteristics are not a valid substitute for personal preference.
Lastly, Boyne looks at twelve studies that attempt to link the self-interests of local government managers and their workforces with the likelihood of contracting out services. The hypothesis here is that public employees resist outsourcing in order to protect their positions within the community at the expense of the public''''s best interests. The variables examine include 1) Ratio of Public Staff to Local Population, 2) Level of Unionization, 3) Level of Wages and 4) Structure of Municipal Government - some believe that whether a municipal government is council-manager or mayor-council plays a role in whether services are contracted out. There is some evidence that a large public employee base will in fact reduce contracting of services. Results for local unionization effects are mixed but that too may have a negative effect. High labor cost is seen as potentially having a positive effect on outsourcing, while mayor-council governments are shown to embrace privatization over council-manager structures. Boyne remains skeptical of these results due to what he sees as possibly erroneous interpretation of the causal effects of the variables used. Once again, he points out that additional extraneous effects of other variables associated with the predictor variables used in the studies may lead to false conclusions.
